CONTACT ENERGY LIMITED V THE ATTORNEY-GENERAL CA CA86/02

CONTACT ENERGY LIMITED V THE ATTORNEY-GENERAL CA CA86/02 4 August 2003IN THE COURT OF APPEAL OF NEW ZEALAND CA86/02BETWEEN CONTACT ENERGY LIMITED Appellant AND THE ATTORNEY-GENERAL Respondent Hearing: 10 March 2003 Coram: Gault P Blanchard J Anderson J Appearances: R A Dobson QC, C S Chapman and M J Bunting for Appellant J S Kos, C J Mathieson and E Fitzgerald for Respondent Judgment: 4 August 2003JUDGMENT OF THE COURT DELIVERED BY ANDERSON J Background of the appeal[1] This is an appeal from a High Court decision about pricing provisions in a contract under which Contact Energy Ltd ("Contact"), as the successor to Electricity Corporation of New Zealand ("ECNZ"), buys Maui Gas from the Crown. The contract was entered into between ECNZ and the Crown on 5 July 1990 and Contact succeeded ECNZ as a party in 1996. The term of the contract is not finite but, as may be inferred from cl 3.1.1, it was expected to continue until 27 June 2009. Its commencement date is stipulated as 30 June 1989. [2] The agreement provides for Contact to pay the Crown on a basis which includes a profit element referred to in the agreement as "the Crown Margin". The relevant terms of the contract, with formulae for the calculation of price including2 the Crown Margin, are set out in the High Court judgment in the following uncontentious terms:[3] Under clause 9.3.1 of the contract Contact is required to pay the Crown a profit margin on the gas it purchases (the Crown Margin), the amount of which is calculated by prescribed formula and is dependent on the gas quantities actually taken and the gas quantities specified in the contract. Clause 9.3.1 provides: In each year ECNZ shall pay the aggregate amount of Crown Margin calculated in respect of the Base Gas for that Year whether or not: (a) ECNZ has actually accepted and taken that quantity of Gas; (b) the Crown was for any reason (including Force Majeure) unable to deliver that quantity of Gas; or (c) ECNZ was for any reason (including Force Majeure) unable to accept and take that quantity of Gas. [4] The prices used to determine the total amount payable under the contract in each year are the Maui Gas Price, the Crown Margin and the Energy Resources Levy ("ERL"). The Maui Gas Price and the ERL are of no consequence in this case: only the calculation of the Crown Margin, essentially from 1 April 2000 onwards, is in issue. [5] The definition of "Crown Margin" is set out in clause 1.2 of the contract. It means: 'Crown Margin' means, at any time, an amount for a GJ [a gigajoule] of gas equal to: (a) NZD2.225 as adjusted by application of clause 9.5; less (b) The aggregate of the Maui Gas Price and NZD0.45 [ie the ERL], in each case, for each GJ of gas. [7] Clause 9.5.1 contains the formula for adjusting the amount of the Crown Margin payable by Contact for each GJ of Base Gas over time. It is as follows: On each successive Adjustment Date the Crown Margin payable for each GJ of Base Gas under Clause 9.3.1 shall be adjusted by: (a) multiplying an amount equal to NZD2.225 by the quotient obtained by dividing PPI by the Initial PPI; and then (b) deducting from the product an amount equal to the Maui Gas Price and NZD0.45. [8] This is conveniently stated as:3NZD2.225 x PPI – NZD0.45 – Maui Gas Price = Crown Margin Initial PPI [9] Only the first part of the above equation is in issue. It comprises: NZD2.225 being the base price of Gas for escalation purposes (set at June 1989) and escalated by the PPI over the initial PPI, with the price adjustment dates set at six monthly intervals on 1 April and 1 October respectively. The margin calculated on each of those dates is then charged for the six month period following. [10] The definitions of "PPI" and "Initial PPI" are set out in clause 1.2 of the contract. The "PPI" means: On any Adjustment Day, the index number shown in the index for the last quarter in respect of which the index has been published and publicly released that ended no less than 3 months prior to such Adjustment Day: for this purpose the index is the Producers Price Index Inputs (all Industries) as from time to time published as table PPIQ.SI9 in the Key Statistics published by the Department of Statistics or the equivalent Government Department or other authorised agency required to prepare and publish such index or its equivalent and if that index is no longer published then such other index or its equivalent as may be agreed by and between ECNZ and the Crown and in default of agreement as to such index within 30 days of either party submitting an index to the other then such index as shall be nominated by the President for the time being of the New Zealand Society of Accountants upon the reference to him by either party. [11] The "Initial PPI" is fixed at 1,600, which was the PPI for the quarter ended 30 June 1989 (Base: the quarter ended 31 December 1982 = 1,000).[3] The obvious purpose of including an index in the price formulae is to maintain the June 1989 value of the Crown Margin by modifying it in terms of a particular measurement of monetary inflation. Against the possibility that the selected index might cease to be usable, the parties stipulated a mechanism for substitution as indicated in Cl 1.2. It is common ground that any substituted index would have to be scaled in a manner that maintained the relativities of the values in the index which had been replaced because of the nature of indexes and the parties' adoption of a pricing formula involving comparison with the Initial PPI. But the Crown contends that as well as having to be scaled, a substituted index would have to be linked with SI9. The terms "scaling" and "linking" are discussed later in this judgment. [4] In 1995 Statistics New Zealand, the present name for the Department of Statistics, decided to redevelop the entire Producers Price Index by replacing the4 New Zealand Standard Industrial Classification (NZSIC) with the Australian and New Zealand Industrial Classification 1993 (ANZSIC). In December 1995 it publicly notified the introduction of an interim index, called SAI9, pending redevelopment. To this end it ceased publishing SI9 in Key Statistics after August 1996 at which time it published SI9 in relation to the quarter ending 31 March 1996. It did, however, continue to calculate SI9 and provide it to people who asked for it, until the quarter ending 30 September 1999. [5] The interim index, SAI9, was first published in Key Statistics in September 1996 and in respect of the quarter ending 31 March 1996. It was last published in December 1998 in respect of the quarter ending 31 March 1998. [6] In December 1997 Statistics New Zealand notified its intention to replace the interim index SAI9 with a fully redeveloped index named SN9. Such index was released in August 1998 and first published in Key Statistics in January 1999 for the quarters ending 30 June 1998 and 30 September 1998. [7] Statistics New Zealand also published a backdated series of values in SN9. It reflected the scaled values of SI9 at December 1995 and SAI9 at December 1997 and extended backwards beyond the commencement date for the purposes of the contract in issue.Scaling and linking[8] An index measures changes in mathematical values relative to a reference known as an "expression base". The attribution of value for an expression base is arbitrary, as it may be since the function of an index is to measure subsequent values relative to the expression base and not to an extraneous reference. An expression base value of 100 or 1,000 is often considered expedient. The indexes with which we are concerned measure changes in the sum of the prices of a selection of goods and services over the course of time. Index SI9 had an expression base of December 1982 with an attributed value of 1,000. As at 30 June 1989, the commencement date for the purposes of the contract in issue, the index value of SI9 was 1,600.5 [9] An index with a certain choice of factors for measuring variations in value, such as the selection of specific goods and services the sum of whose prices is to be measured over time, will be different from any other index which does not have identical factors. An index developed for certain purposes may have its utility reduced by the developing obsolescence of the measurement factors, or their relative weighting. An example given by counsel for the appellant points to the greater significance in a modern commercial context of laser printer accessories than, say, typewriter ribbons which may have been included in an ageing basket of goods and services. [10] If the expression base of an index were to be changed in accordance with a mathematical function, the relativity of all the values in the index will remain constant if they are modified by the same function. Thus, for example, if the value of an expression base of 1,000 were doubled to 2,000 then the relativities of all the values in the index will remain constant if they also are multiplied by two. Such a process is called "scaling" or "rebasing". [11] It is possible to join together, or link, indexes if one is scaled so that it has the same value as the other at the selected linking point. Suppose that Index A has a value at a certain point of 2x and Index B has a value at that point of x; the indexes could be linked at that point if the values of Index B are scaled so that x = 2x. Up until the point Index A can apply and from the point Index B can apply. The relative values are the same for each index at the selected point. This process of linking is sometimes called "splicing". Where it occurs the result may be considered a composite index, being neither wholly one nor wholly the other even though there is a seamless transition between them; but from the point of transition forward the relative values are determined by the index to which the anterior was spliced. [12] In this case the relevant indexes measure values for time periods marked by the ending of annual quarters. For two indexes to be spliced they must have some measure of overlap. This means that splicing could occur at the first or at the last point of correlation, presenting the respective options of splicing backwards or splicing forwards. The point of splicing can be, and in this case is, important because values after splicing will be relative to the value at the point of splicing.6 [13] Suppose that during the period of overlap the relative values of Index A increased and the relative values of Index B did not. Splicing at the first point would not capture that relative increase, whereas splicing at the second point would. [14] In the present matter it is relevant that significant divergence in values occurred between SI9 and SAI9 in the quarter ending March 1998, and between SI9 and SN9 from March 1998 to September 1999, the values of SI9 being greater in each case. The outcome for which the Crown contends in this litigation would give it the benefit of those greater values; conversely, Contact's case would give it the benefit of the lower values. [15] As previously mentioned, Statistics New Zealand calculated and provided SI9 up until the quarter ending 30 September 1999. Paragraph 20 of an agreed statement of facts acknowledges that this amounted to publication by Statistics New Zealand, albeit not in Key Statistics after September 1996. [16] The parties continued to use SI9 until June 1999, with the Crown invoicing and Contact paying on that basis. The last publication of SI9 occurred in November 1999, between the price adjustment dates of 1 October that year and 1 April 2000. Its values referred to the quarter ending 30 September 1999. [17] Negotiations between the parties as to a new index began with a suggested approach by the Crown to Contact in a letter dated 26 April 2000 from Mr P McKenzie to Mr A Love. There was an exchange of correspondence for some three months culminating in: (a) a letter dated 31 July 2000 from Mr McKenzie recording, amongst other things, that:Matters not in dispute are: • That the index that should apply in the future would be PPIQ.SN9 index as that index is described by Statistics New Zealand as the All Industries PPI Inputs index(b) a letter in reply from Mr Love dated 10 August 2000 advising, amongst other things,7We confirm that PPIQ.SN9 is the applicable substitute index.[18] Although of the same mind as to the appropriate index, the parties could not agree upon its application. The Crown insisted that September 1999 was the linking date anticipated by the contract and Contact maintained that the Crown Margin should be adjusted in accordance with changes in SN9 from the quarter ending 30 June 1989. Goddard J described Contact's approach, which Contact maintains is correct, as "simple retrospective substitution of SN9 for SI9 as far back as 30 June 1989". [19] In his letter of 10 August, Mr Love considered that Contact had been overcharged through being invoiced between 1 October 1996 and 31 March 2000 on the basis of SI9 rather than SAI9 from October 1996 until December 1997 and thereafter on the basis of SN9. He also took issue whether SI9 had been published in Key Statistics for the quarter September 1999. [20] On 8 September 2000 the Crown's solicitors wrote to Contact, describing the issue between the parties in the following terms:[3] The issue between the parties, therefore, is when "that index (ie PPIQ.SI9) [was] no longer published". The Crown contends that that was as from 1 October 1998; Contact, as from 1 April 1996. Contact's position is dependent on PPIQ.SI9 having to continue to be published in "Key Statistics" in order to continue to govern the position, which assertion the Crown rejects.[21] On 20 September 2000 the Crown advised Contact of its intention to invoice on a basis linking SN9 in respect of the September quarter 1999. The following day Contact wrote to the President of the Institute of Chartered Accountants seeking to invoke the dispute resolution provisions in the definition of "PPI" noted in paragraph [10] of the High Court judgment reproduced in paragraph [2] of this judgment. The Crown pre-empted that intended course by informing the President that there was no dispute to submit because the substitute index had been agreed. [22] Ultimately the Crown sought to surmount the impasse with litigation. It sued in the High Court for damages in the sum of $2,253,185.90, being the difference between the amounts claimed down to November 2001 on the basis of its assertion8 that SN9 applied from September 1999, and the amounts actually paid by Contact in reliance on that party's view. The Crown also sought interest in terms of contractual provisions, costs and declarations in the following forms:• A declaration that SI9 continued to be published for the purposes of the Contract until the September 1999 quarter.• A declaration that on the Adjustment Date on 1 October 1999 the relevant PPI for the purposes of calculation of adjustment of the Crown Margin under the Contract was SI9.• A declaration that on the Adjustment Date on 1 April 2000 and on all subsequent Adjustment Dates the relevant PPI for the purposes of calculation of the adjustment of the Crown Margin under the Contract was SN9, rebased so that it equals SI9 for the quarter ending 30 September 1999.High Court judgment[23] To facilitate the trial the parties agreed upon a statement of issues as well as a statement of agreed facts. The terms of the issues and Goddard J's answers to the same are reproduced immediately below:1. Does the definition of PPI in clause 1.2 of the Contract require that SI9 continue to be used for the purpose of price adjustment until it is no longer published, or does it require that SI9 continue to be used for price adjustment only so long as it is published in Key Statistics? No. The reference to Key Statistics is descriptive rather than prescriptive.2. Irrespective of the answer to (1) above, did the parties agree to continue to use SI9 notwithstanding the cessation of its publication in Key Statistics. If so, what were the terms of such agreement?Yes. The parties agreed orally in October 1996 to continue to use SI9, rather than move to SAI9. This oral agreement could be construed as an agreement for the purposes of clause 1.2 of the Contract. In any event, even if the parties did not reach such an agreement, they nevertheless continued to use SI9 until it ceased to be published in any form. Their ongoing conduct in this regard amounted to an unconditional agreement which neither party sought to bring to an end, but which necessarily expired upon SI9 ceasing to be published at all.93. After SI9 ceased to be published, the parties agreed to use SN9 for the purpose of calculating the Crown Margin as from and including the Adjustment Date of 1 April 2000. In this context: (a) On a true construction of the Contract, does the definition of PPI in clause 1.2 of the Contract require SI9 and SN9 to be linked to one another (as the Crown contends)? or (b) Is a term to be implied in the definition of PPI in clause 1.2 of the Contract requiring SI9 and SN9 to be linked to one another (as the Crown contends in the alternative)? or (c) On a true construction of the Contract is SN9 to be substituted for SI9 as the PPI, so that SN9 thereby became the "PPI" as defined in clause 1.2 of the Contract (as Contact contends)?As to (a) and (b): Yes. This is the accepted and statistically correct means of transitioning from one index to another, which the parties must have intended when they entered into the contract. Furthermore, "PPI" by definition provides for any replacement index to be linked to SI9. The wording of the definition of PPI should therefore be interpreted in accordance with such an intention; or alternatively a term implied to reflect such an intention. As to (c): No. Such an approach involves a change to the contractual definition of Initial PPI, from 1600 to 863, which is not provided for under the contract and there has been no agreement by the parties to depart from the terms of the contract. Simply substituting SN9 for SI9 would lead to consequences not intended by the parties: for example, it would have the effect of changing the contractual definition of "Initial PPI"; re-writing history; and produced a "false quarter".4. If the approach taken in paragraph 3(a) or 3(b) above is correct: (a) At what date should the two indexes be linked?At September 1999, the last quarter for which SI9 was published and used by the parties. As linking preserves the movements in the "contract" index up to the link date, September 1999 is the only logical point at which to link.(b) Does the "Initial PPI" index number remain 1600 as defined in clause 1.2 (as the Crown contends)? If not, what does the initial index number become?10Yes. Under 3(a) and 3(b), the Initial PPI does not change. As the correct approach is to link (on a linking forward basis) as at September 1999, 1600 remains constant.5. If the approach taken in 3(c) above is correct: (a) Does the "Initial PPI" index number remain 1600 as defined in clause 1.2?No it does not.(b) If not, what does the Initial PPI index number become?863, being (backdated) SN9 at 30 June 1989.6. What is the PPI index number for each Adjustment Date calculation?(a) for 1 April 2000: 2925 (b) for 1 October 2000: 1974 (c) for 1 April 2001: 2121 (d) for 1 October 2001: 2144 (e) for each successive Adjustment Date, the appropriate SN9 index number scaled by a link factor of 1897 10147. Using the Initial PPI index number and the PPI index number determined by the Court in respect of the relevant Adjustment Dates, what amounts did Contact owe to the Crown on each of the invoices dates listed in paragraph 30 of the second amended statement of claim?(a) For the 1 April 2000 Adjustment Date and the six successive months thereafter: $3,222,835.000 (ex GST). (b) For the 1 October 2000 Adjustment Date and the six successive months thereafter: $3,213,7700.83 (ex GST). (c) For the 1 April 2001 Adjustment Date and the five successive months thereafter: $3,691,200.83 (ex GST). (d) For the 1 October Adjustment Date and the month of October 2001: $2,434,688.74 (ex GST).8. What, if any, debt owed by Contact to the Crown remains unpaid?The sum of $2,253,185.90 (incl GST) (ex interest).11 [24] Pursuant to those findings Goddard J entered judgment for the Crown for the sum of $2,253,185.90 (inclusive of GST) together with interest, costs and declarations in the terms sought by the Crown. [25] The Judge's essential reasons can be stated quite briefly. She answered the first issue on the basis of the interpretation of the PPI definition clause; and the second question on the basis of the evidence, which included much documentation although no testimony except by witnesses for the Crown. [26] On the third issue Goddard J held that in expressly providing for the nominated index to be replaced, cl 1.2 inherently provides for any replacement index to be linked at the point at which the transition to the replacement index is made. She considered such a view was supported by the expert evidence that linking is so integral to a transition from one index to another that it would be otiose expressly to spell out the need for it. The only time there could be a linking was when the parties ceased to use the original index. In this case the parties ceased to use the movements in SI9 as at 30 September 1999 and accordingly that was the time when the linking had to occur. She held that "to link at any other point would, on the evidence of the experts, fail to produce one continuous linked index spanning the entire period of the contract and would thus fail to preserve all price relatives up to the point of transition and carry them into the replacement index". [27] The Judge considered it axiomatic that the necessary linking in the present case must be forward from the actual point of transition as that was the only way in which a continuous sequence of inflationary movements, referenced back to the initial contract price, could be achieved. She considered that Contact's approach in advocating simple substitution of SN9 without linking it to SI9 from the point of transition rested on an assumption that the only objective in using a price index is to ascertain a value today of a 30 June 1989 price. However, in the Judge's view, the purpose of a contract index is to measure price change over time relative to the initial contract price. [28] Goddard J had regard also to the evidence of an expert statistician, Dr P J Thomson, in respect of what was described as a "false quarter". Dr Thomson12 expressed the opinion that a simple substitution of SN9 for SI9 from 1 April 2000 onwards would dislocate vital continuity and override the historical position. In his view the result would be a drop of 1.07% effectively creating a false quarter of inflation not reflected in any index including SI9 or SN9, not reflecting any real price movement at all and negating the whole point of having a contract index. [29] The Judge held that even if she were wrong in her conclusion that the provision in cl 1.2 for a transition to a replacement index is, by definition, a provision to link at the point of transition (forward), such a provision must at the very least be an implied term of the contract. This is because linking is so clearly inherent in a transition from one index to another that it must be implicit in any contractual term providing for transition. Such an implied term would be reasonable and equitable; necessary to give business efficacy to the contract; so obvious that it goes without saying; capable of clear expression; and would not contradict any express term of the contract. It thus satisfied the conditions identified in BP Refinery (Westernport) Pty Ltd v Shire of Hastings (1977) 16 ALR 363. [30] A further and related issue for the Judge's determination was whether the Crown was estopped from denying that the parties had agreed to use SN9 as a replacement index, to the intent that SN9 would be retrospectively substituted for SI9 back to 30 June 1989, with SN9 rebased to equal the initial PPI. The Judge's rejection of the aptness of that approach has been discussed earlier in this judgment. With specific reference to estoppel, she held that SN9 rebased to equal SI9 for the quarter ending 30 September 1999 remained SN9 rather than a "composite custom- made index" as Contact had submitted. The defence of estoppel was held to fail accordingly.Appellant's arguments[31] At the hearing of this appeal Contact abandoned its challenge to the Judge's conclusions in respect of the first stated issue, which was concerned with whether the definition of PPI required that SI9 continue to be used for the purpose of price adjustment until it is no longer published or whether it requires that index to be used only as long as it is published in Key Statistics. The parties had however filed13 written submissions on that issue and, without traversing them we express the view that Goddard J's conclusion was entirely correct. [32] As to the second issue, counsel submitted that the High Court erred in finding an oral agreement to continue using SI9 because no such agreement was pleaded. Further, the parties did not at any time agree to continue using SI9 notwithstanding the cessation of publication in Key Statistics; and if they did so agree, it was not a term of such agreement that they were obliged to use it until Statistics New Zealand ceased to make it available in any form. Contact argued that the correspondence adverted to by the Crown in its statement of claim does not disclose any such agreement and, if the agreement were oral, all the Crown can point to is a brief telephone discussion between Mr P McKenzie and Mr Love, the effect of which cannot be an agreement to use SI9 without limitation. Given the appellant's abandonment of its challenge in respect of the first issue we think the second issue is of limited relevance. If the contract mandated use of the particular index until publication ceased, an agreement to continue to use it is of no significance. [33] The substance of Contact's submissions concern the issues relating to the application of SN9. The essential themes of the argument are: (a) A replacement index, be it agreed by the parties or nominated by the President of the New Zealand Institute of Chartered Accountants, (the President) must be an index which provides both index numbers required by the contract; these are the initial PPI, value 1,600 at June 1989, and the PPI number for the relevant adjustment date. (b) There is no provision in the PPI definition or elsewhere in the contract for linking. A replacement index can always be agreed or nominated which the parties can use without linking. (c) The absence of any provision in the contract prescribing how an index is to be constructed shows an anticipation that the parties will14 contend for replacement indexes that are already completely constructed. (d) This approach is straight-forward and needs no implication of terms. It works adequately without requiring such an implication and leaves little scope for dispute other than what index should be agreed or nominated. [34] It was submitted for Contact that the High Court judgment was wholly reliant on a conclusion that the contract required the parties to maintain a continuous linear sequence, a conclusion which required use of the Crown's approach because Contact's "would dislocate vital continuity and override the historical position". The Judge had said, "The important aspect was to ensure that price changes moved forward in a continuous linear sequence from the fixed initial reference point of the initial PPI of 1600"; and "the contractual objective of maintaining all price relatives for the term of the contract"; and transition from one index to another had to be "achieved without affecting the continuity of linear sequence so vital to preservation of relativity in price measurement". [35] Counsel submitted that if the Judge was wrong about the contract requiring the parties to maintain a continuous linear sequence then her conclusions lose their foundation. And, it was argued, she was wrong so to find because she erred in thinking there was evidence to that effect and, there being no express term to that effect, erred in finding that a term had to be implied into the contract to that effect. [36] Counsel argued that it was not legitimate to imply a term for linking as that was not necessary to give business efficacy to the contract, nor was it so obvious that it goes without saying. So long as there was agreement or nomination of an index that satisfied the requirements of providing the same expression base, an initial PPI number and a PPI number for the adjustment date, then the contract functions effectively on its plain terms. It was submitted that all three indexes under consideration in this proceeding fulfil the criteria and there is accordingly no need to imply a linking provision.15 [37] It was submitted that the Crown's experts accepted that it would have been wholly open to the parties to deal with the question of transition in some other way than by a linking provision and, in any event, the term sought to be implied has a high potential for encouraging disputes. [38] The judgment under appeal was also criticised on the basis that 30 September 1999 was not the date when the parties ceased to use movements in SI9, the relevant date having been 30 June 1999. This was said to be an error because the contract uses only the index numbers for the December and June quarters. [39] Contact also submitted that Goddard J was wrong to conclude that its argument would result in the Court having to rewrite the definition of the initial PPI to substitute a different value. That conclusion was contrary to the Court's earlier finding that "it is and was common ground between the parties that SN9 (or any other replacement index) would have to be scaled before it could be compared to the contractually fixed initial PPI of 1600". [40] Contact submitted that the combination of the expert's evidence and the Judge's reasoning are essentially circular, the evidence depending upon assumptions about what the contract means and the Judge relying on that evidence to determine what the contract means. Once the experts acknowledged that the parties could have agreed to what Contact argued they did agree upon, their role was over. However, they went on to give evidence as to how to interpret the contract and such evidence should have been ignored by the Court. [41] Contact argued that the Crown was estopped from denying that SN9 is to be used, it being crucial for such argument that the linked index propounded by the Crown is not SN9 whereas the unlinked but rebased initial PPI index propounded by Contact is. [42] Contact criticised the Crown's proposal on a number of grounds. It was said that a continuous linear sequence was not a contractual objective, nor was statistical purity. Contact's position achieved a commercially rational or fair outcome because it promoted the agreed purposes of the contract rather than an abstract notion of16 statistical purity. The purpose of the contractual provisions was to achieve current value of the June 1989 price. Counsel pointed to evidence given by Mr P McKenzie in examination in chief and cross-examination to the effect that the base price for gas was to be maintained over the life of the contract in real terms; that a critical requirement of the index chosen by the parties included that it should be a fair reflection of real price changes in the inputs (all industries) sector; and that to achieve the results in real terms one would have to use the official PPI all the way through and have it linked as Statistics New Zealand linked it. In counsel's submission, had that been done, then the experts were quite clear that the way in which prices should be determined from 1 April 2000 is by using the SN9 series scaled so the SN9 index number for the quarter ending June 1989 is 1600. This is what Contact says should be done but the Crown contended, illogically, it was said by Contact, for other than an official index throughout, which would produce a result that did not accord with the parties' contractual objective. Further, the parties must have intended when they entered into the contract that the effect of any departures from the official index would be minimised. [43] Contact also argued that the question of linking was essentially one of construction and that Goddard J was wrong to find as a matter of construction first, that linking is a term of the contract and second, that the linking has to be at the point of transition. It was submitted that those determinations were arrived at, not as a result of construing the contract in a conventional sense but in reliance on the evidence of Crown witnesses as to what they thought the contract meant or should mean. Such a method is inappropriate because the meaning of a contract is to be found by an analysis of its terms, in the commercial context in which it was concluded, and not by relying on conjecture of non-commercial witnesses with a technical interest not shared by the parties themselves, and with the benefit of hindsight. [44] A consequence of that approach, it was submitted, is that should the reference to the President be invoked, the President would not be nominating an index to replace SI9 but would be nominating only part of an index with a composite index resulting. But the contract does not provide for the President to nominate only part of an index and the restricted process would convert a factual issue within the17 expert's competence into one partly to be resolved by the expert and partly to be resolved by the implication of a term. This approach would prevent the President adjudicating on the economic and commercial merits of competing indexes and would turn a simple expert determination into a convoluted and unnecessary combination of factual and legal issues. This is inconsistent with a contractual purpose of minimising the possibility of future disputes and ignores the fact that the contract adjusts the Crown Margin by reference to a base price and base index rather than by reference to the price and index used six months previously. [45] Finally, Contact submitted that should it be found, notwithstanding the appellant's arguments, that the parties have not in fact agreed to a replacement index, then, because the President has not nominated an index, this Court should determine the appropriate index on a factual determination of which index best serves the agreed purpose of the relevant provisions. That was not an issue before the High Court. We do not set out the arguments in support of that proposition because it is quite untenable that this Court should deal with that matter as if the appeal were an ad hoc arbitration. There are only two ways in which, under the contract, a necessary index can be imposed on the parties. They must either agree or the President must nominate an index.Respondent's arguments[46] Counsel for the Crown submitted that in both the Crown's approach and Contact's there is a linking of indexes, the difference being as to the chronological point at which this occurs. The Crown contended for September 1999 when SI9 ceased to exist. Contact's approach inherently links at December 1997, at which date SN9 was linked to SAI9 which itself had been linked to SI9 at December 1995. [47] The Crown characterised Contact's approach as purely opportunistic and producing a result which did not work or make sense because: (a) It produced the false quarter, as the Judge had found (see para [27] of this judgment).18 (b) It mandates an index which has a value for June 1989. (c) It mandates, in fact, SN9 because that is the only index which has a value of 1,000 in December 1982 and 1,600 in June 1989. (d) It makes dissent inevitable because of its inherent potential advantages and disadvantages. (e) It is statistically unsound. [48] On the other hand, it was submitted, the Crown's approach provides a seamless transition permitting any index to be used. This would encourage agreement because only future movements in value will be relevant and because it is in any event the normal and usual means of transition. [49] The dispute in the present case is not over the choice of index, however, but over the method of its implementation. It was submitted that Goddard J was correct to find, as a matter of construction either expressly or by implication, that the contract required a method of transition to the new index by linking at the time the old index ceased to be used. Her choice of a linking date in September 1999 was correct because that is what cl 1.2 of the contract provides. In any event, the Judge was correct to find there was an agreement to continue using SI9. There was uncontradicted evidence to support her finding in this respect. [50] It was submitted on behalf of the Crown that there could be no question of estoppel because there was no evidence of the Crown ever agreeing to the position asserted by Contact. The Crown and Contact each accepted that SN9 was the index to be used but the Crown had never represented that it would agree to retrospective substitution, insisting instead on linking forward.Discussion19 [51] We agree with Goddard J's conclusions on all the defined issues although our reasons place rather more emphasis on the necessity to imply a term in the contract. But we concur with her view that the case is essentially one of construction. [52] The contract must, of course, be construed, not in light of the fortuity of the development of SN9, but in terms of objectively assessed contractual intent at the time the contract was entered into in July 1990. Clause 3.1.1 indicates an expectation that the term would continue for some 20 years, given the 12 months notice required upon the expiration of a 30 year period on 27 June 2009. The subject matter of the contract was the supply of potentially vast quantities of gas for monetary consideration which would be very substantial. The parties envisaged that the index by reference to which the Crown Margin was to be determined might cease to be usable for that purpose during the life of the contract and that they might not be able to agree upon a replacement index. Should they not, the President was to nominate an index but could do no more than that. The President could not make a determination that the nominated index was to be linked or, if it were to be linked, whether it should be linked backwards or forwards or at what time it should be linked. Nor are there any criteria stipulated for the selection by the President. But, given the commercial nature and significant scope of the contract, the parties must be taken to have assumed that a nominated index would have a rational relevance to the future maintenance of the value of the Crown Margin in terms of its formula for assessment. [53] To have a rational relevance an index would obviously need to be current and continuing, not historical and extinct, at the time of substitution. What may have happened in the past in relation to the pricing of the Crown Margin would not require revisiting in terms of a replacement index. The future maintenance of the relative value of the Crown Margin at the time of substitution would be achieved by reference to the future values of a replacement index. It is quite plain to our mind that the intended purpose of a replacement index was to apply its future values, from the time of the substitution, to the relative value of the Crown Margin fixed by reference to the SI9, at the point of substitution. Although the result would be a composite index from the commencement of the contract, there would be a single series of index values from the date of transition. The parties obviously intended20 that a replacement index would be nominated for the purpose of the application of the future values of that index. [54] It could not be assumed at the time the contract was entered into that any other rationally relevant index would have a value for June 1989 and the parties must be taken to have envisaged the possibility that the only rationally relevant indexes able to be invoked in substitution would not have a value for June 1989. In that event, there would have to be a scaling and linking of any index which might be selected because if that were not done the new index would have no logical relevance to what it was replacing. Suppose SI9 had a value at a hypothetical point of future substitution of, say, 2,500 and a nominated index at that time had a value of 1,000, it would be absurd to contemplate future prices commencing with a PPI of 1,000 or, ignoring the value reached at the point of time when SI9 could no longer be used, to commence the new index values at an earlier time. Commercial reality could be maintained only by scaling and linking. Since the parties must have been concerned to ascertain future prices by reference to future values rather than past prices by reference to past values, the only logical method of linking would be to link forward. [55] The evidence given in the High Court showed, as Goddard J found, that "linking is so integral to a transition from one index to another that it must be otiose to expressly spell out the need for it". But as well as there being an evidential basis for the inference that linking is so inherent in the function of transition, reason dictates that this must have been the unstated intention of the parties because they must have assumed a rational, not an absurd, response to a future unavailability of SI9. When one approaches the matter in terms of 1990 intention as to future possibilities rather than 1999 exploitation of actualities it becomes obvious that there must be implied into the contract a term to the following effect, as a final sentence to the definition of PPI in cl 1.2:Such index, agreed or nominated as the case may be, shall be linked to SI9, forward[56] Such term must be implied because:21 1. It is reasonable and equitable in that it maintains the relative value of the Crown Margin, assessed in accordance with the formula at the time SI9 became no longer usable. It is also reasonable because it mandates a conventional process of transition from one index to another. It is equitable because the parties plainly intended that the Crown's entitlement and Contact's liability would be principled, not arbitrary and the term neutralises the advantage or disadvantage, to one party or the other, of arbitrary values in a replacement index. 2. It is necessary to give business efficacy to the contract because the mere nomination of an index without regard to its chronological relevance and relativity in terms of value would produce an absurdity. The fact that a decade after the contractual intent was formed there happened to be a similar index with similar values including for June 1989, cannot inform that prior intent. 3. It is so obvious it goes without saying, both because linking is inherent in index transition and because the parties were concerned with prospective, not historical values. 4. It is capable of clear expression. 5. It does not contradict any of the terms of the contract. [57] The time for linking must be September 1999 because of the terms of the definition of PPI and the facts of the case. The relevant adjustment day in terms of the contractual definition is 1 April 2000. The last quarter which was published and publicly released was September 1999. That quarter ended no less than three months prior to the Adjustment Day of 1 April 2000. The relevant quarter could not be June 1999, as the appellant contends, because that was not the last quarter in terms of the contractual definition but the penultimate quarter. It is the case that Goddard J refers to the last quarter for which SI9 was published and used by the parties but we think it obvious from the context that she meant "required to be used" rather than merely "used". In any event, Contact is plainly caught by the terms of the definition and the facts of the case. [58] We turn to consider the appellant's arguments in light of our view as to what the agreement envisages. We do not accept that a replacement index must be one which provides values for 31 December 1982 and June 1989. No such requirement is expressed or implied in the agreement. Linking is to be implied for the reasons we22 have indicated. Whether the parties might or might not, in the event, contend for fully constructed indexes is not relevant to contractual intent. [59] The appellant's argument that the agreement did not mandate a continuous linear sequence, contrary to the Judge's view, could be right only if the contract mandated that the parties disregard the relative value of the Crown Margin in terms of the formula which had been agreed and that they should start from scratch to determine a present value at the time of transition by reference to another index. And we have rejected that possible intent which would have to assume the future existence of a rationally relevant index which had a value for December 1989. [60] We turn now to the appellant's argument that the Crown is estopped from denying that SN9, as a whole index, is what the parties have agreed to apply, without scaling and linking at the transition point. The appellant's argument ignores the contractual context and exploits the fortuity of some coincidental historical values. [61] Given the term which must be implied the index identified by agreement must be linked forward at the point of transition. [62] Goddard J found, contrary to the appellant's argument before her, that SN9, rebased to equal SI9 for the quarter ending 30 September 1999, remains SN9 and is not some "composite custom-made index" not agreed to by the parties as the replacement index. She rejected Contact's argument that SN9 rebased to equal SI9 at 30 September 1999 is a composite custom-made index and not SN9 at all. On that basis she held that the affirmative defence of estoppel fails. [63] We think Goddard J was right to reject the estoppel argument. If such an argument were to prevail in respect of SN9 then it would have prevailed if any other index happened to have been agreed upon. The characteristics of SN9 do not inform the parties' intention at the time the contract was entered into and the appellant's argument assumes that the parties intended that the nomination, by agreement or by the President, of any future replacement index would not involve rebasing and linking but mere substitution of values. We have rejected that approach as untenable and held that the parties intended any selected index to be rebased and linked23 forward. It may be said that looking at the whole life of the contract the Crown Margin will have been determined by a composite index if there is a substitution but we do not accept that this is inconsistent with the contractual intent.Conclusion[64] For the above reasons we are of the view that Goddard J answered all the agreed issues correctly and that the appeal must accordingly be dismissed with costs, which we fix at $6,000 together with disbursements including the reasonable travelling and accommodation expenses of two counsel for the respondent.Solicitors: Buddle Findlay, Wellington for Appellant Crown Law Office, Wellington for Respondent